%% Grandfathering Project
%% Robustness regression table
%% Compiled on 2024-07-04 06:22:29 using <5_regs_robust.R>
%% Written by Jack Gregory

\begin{center}
\begin{singlespace}
\begin{scriptsize}

\begin{longtable}[c]{@{\extracolsep{0.0ex}}l*{4}{D{.}{.}{-2}}@{\extracolsep{5.0ex}}}
	\caption{Robustness regression results}
	\label{tbl:reg_robust}

\\
\hline\hline \\ [-1.2ex]
 &	\multicolumn{3}{c}{\textit{Non-linear Age}} &			\multicolumn{1}{c}{\textit{Spillover}}	\\\\[-1.8ex] \cmidrule{2-4}\cmidrule{5-5}
 &	\multicolumn{1}{c}{\textit{Utilization}} &	\multicolumn{1}{c}{\textit{Survival}} &	\multicolumn{1}{c}{\textit{Emissions}} &	\multicolumn{1}{c}{\textit{Utilization}}	\\
 &	\multicolumn{1}{c}{(1)} &	\multicolumn{1}{c}{(2)} &	\multicolumn{1}{c}{(3)} &	\multicolumn{1}{c}{(4)}	\\\\[-1.8ex]
\midrule \\
\endfirsthead
\hline\hline \\ [-1.2ex]
 &	\multicolumn{3}{c}{\textit{Non-linear Age}} &			\multicolumn{1}{c}{\textit{Spillover}}	\\\\[-1.8ex] \cmidrule{2-4}\cmidrule{5-5}
 &	\multicolumn{1}{c}{\textit{Utilization}} &	\multicolumn{1}{c}{\textit{Survival}} &	\multicolumn{1}{c}{\textit{Emissions}} &	\multicolumn{1}{c}{\textit{Utilization}}	\\
 &	\multicolumn{1}{c}{(1)} &	\multicolumn{1}{c}{(2)} &	\multicolumn{1}{c}{(3)} &	\multicolumn{1}{c}{(4)}	\\\\[-1.8ex]
\midrule \\
\endhead
GF &	1045.19^{*} &	0.07 &	1.21 &	2447.70^{**} \rule{0pt}{2.6ex} \\
 &	(1.67) &	(0.08) &	(0.99) &	(2.29) \\ [1.0ex]
size &	1956.88^{**} &	2.20^{**} &	0.76 &	2830.89^{**} \\
 &	(2.28) &	(2.33) &	(0.67) &	(2.47) \\ [1.0ex]
GF $\times$ size &	-1496.43^{*} &	-1.33 &	-1.23 &	-2359.55^{**} \\
 &	(-1.81) &	(-1.38) &	(-0.92) &	(-1.99) \\ [1.0ex]
NAAQS &	533.38^{**} &	1.25 &	1.83^{*} &	-945.62^{***} \\
 &	(2.13) &	(1.62) &	(1.78) &	(-3.47) \\ [1.0ex]
GF $\times$ NAAQS &	-1355.95^{***} &	-1.18 &	-2.46^{**} &	 \\
 &	(-3.90) &	(-1.13) &	(-2.03) &	 \\ [1.0ex]
MMBTU &	-76.72 &	-0.02 &	0.09 &	-223.85 \\
 &	(-1.05) &	(-0.07) &	(0.32) &	(-1.20) \\ [1.0ex]
GF $\times$ MMBTU &	-148.91^{*} &	-0.40 &	-1.11^{***} &	-183.24 \\
 &	(-1.75) &	(-1.30) &	(-3.15) &	(-0.91) \\ [1.0ex]
\hline \\ [-1.8ex]
Observations &	\multicolumn{1}{c}{9,230} &	\multicolumn{1}{c}{11,686} &	\multicolumn{1}{c}{9,070} &	\multicolumn{1}{c}{7,982} \\
R$^{2}$ &	\multicolumn{1}{c}{0.323} &	\multicolumn{1}{c}{0.085} &	\multicolumn{1}{c}{0.418} &	\multicolumn{1}{c}{0.309} \\
\hline\hline
\multicolumn{5}{p{10cm}}{\textit{Notes:} Robustness checks based on specification (3) from Table (1), including relevant sample. "Non-linear Age" regressions additionally control for squared age, while the "Spillover" regression uses a sample in which direct interactions between non-grandfathered and grandfathered boilers are limited. Boiler-level clustered standard errors used, with *** p$<$0.01,  ** p$<$0.05,  * p$<$0.10 and \textit{t}-statistics in parentheses.} \\
\\
\end{longtable}

\end{scriptsize}
\end{singlespace}
\end{center}
